Jump to content

Module:Unicode convert

Permanently protected module
From Wikipedia, the free encyclopedia

local p = {}

-- NOTE: all these functions use frame solely for its args member.
-- Modules using them may therefore call them with a fake frame table
-- containing only args.

p.getUTF8 = function (frame)
	local ch = mw.ustring.char(tonumber(frame.args[1] or '0', 16) or 0)
	local bytes = {mw.ustring.byte(ch, 1, -1)}
	local format = ({
		['10'] = '%d',
		dec = '%d'
	})[frame.args['base']] or '%02X'
	for i = 1, #bytes do
		bytes[i] = format:format(bytes[i])
	end
	return table.concat(bytes, ' ')
end

p.getUTF16 = function (frame)
	local codepoint = tonumber(frame.args[1] or '0', 16) or 0
	local format = ({ -- TODO reduce the number of options.
		['10'] = '%d',
		dec = '%d'
	})[frame.args['base']] or '%04X'
	if codepoint <= 0xFFFF then -- NB this also returns lone surrogate characters
		return format:format(codepoint)
	elseif codepoint > 0x10FFFF then -- There are no codepoints above this
		return ''
	end
	codepoint = codepoint - 0x10000
	bit32 = require('bit32')
	return (format .. ' ' .. format):format(
		bit32.rshift(codepoint, 10) + 0xD800,
		bit32.band(codepoint, 0x3FF) + 0xDC00)
end

p.fromUTF8 = function(frame)
	local basein = frame.args['basein'] == 'dec' and 10 or 16
	local format = frame.args['base'] == 'dec' and '%d ' or '%02X '
	local bytes = {}
	for byte in mw.text.gsplit(frame.args[1], '%s') do
		table.insert(bytes, tonumber(byte, basein))
	end
	local chars = {mw.ustring.codepoint(string.char(unpack(bytes)), 1, -1)}
	return format:rep(#chars):sub(1, -2):format(unpack(chars))
end

return p